The numbers don’t match up.
For years psychologists assumed standard questionnaires could measure depression across the board. They couldn’t. Stanisław Czerwiński and his team at the University of Gdańsk proved that two common scales fail when you bring intelligence into the equation. Not just a little fail. They break entirely when comparing people of differing cognitive levels.
The findings, published in the journal Intelligence, suggest this isn’t isolated. If depression scales can’t handle IQ variance other mental health tools probably can’t either.
The Curve No One Expected
Czerwiński started with a hypothesis that sounds intuitive enough until it doesn’t. He guessed that higher intelligence correlates with better mental health… up to a point. After that threshold the relationship flips. The smartest people in the room might be suffering more.
To test this the team looked at data from two massive U.S. surveys. Decades of tracking. Thousands of participants. They used aptitude tests covering math and language as a proxy for IQ. Then they layered on mental health questionnaires covering the usual suspects mood sleep appetite.
The initial data looked right. The curve held. High IQ linked to poorer mental health.
But the scientists didn’t stop there. Good scientists don’t.
The Ruler Made of Putty
They ran validity checks. Specifically they checked for measurement invariance. That’s the technical way of asking “Does a score of 5 mean the same thing to a genius as it does to an average person?”
It doesn’t.
Both scales failed the test. Responses to individual questions reflect different degrees of depression depending on intelligence. Which means the initial finding is trash. Or at least untrustworthy. You can’t compare apples to oranges when your measuring device stretches and shrinks arbitrarily.
Nicole Beaulieu Perez a psychiatric nurse at NYU who wasn’t involved in the study puts it simply.
“Imagine we’re measuring height but our ruler is made of Sililly Putty so the length changes. How can we know how tall these people really are?”
It’s a messy analogy. And accurate.
Why Do Smart People Answer Differently?
The study doesn’t explain the mechanism. Why does intelligence distort these answers? Czerwiński isn’t surprised.
“These questionnaires require interpretation.”
A highly intelligent person might analyze their feelings differently. They might contextualize sleep loss or appetite changes in ways the rigid options don’t capture. They think about their pain differently. This nuance gets lost in multiple-choice bubbles designed for the average responder.
A Broader Measurement Crisis
This breaks current research. Previous studies comparing groups without accounting for these intelligence differences likely drew flawed conclusions. Even clinical screenings in doctor’s offices are suspect. If your ruler is crooked you can’t trust the height.
Perez notes this is likely a systemic issue. She recently reviewed evidence that these scales work consistently across gender and culture. It’s inadequate. Depression is one of the most studied constructs in psychology. And we still can’t measure it properly.
The path forward isn’t more paper.
Researchers are pushing for digital tracking. Objective data like actual sleep duration rather than self-reported restlessness. Experience sampling where users log feelings at random intervals captures reality better than retrospective interpretation. It’s harder to lie to a ping on your phone at 2 AM than to a survey you fill out after the fact.
Czerwiński says the problem is almost certainly wider than depression. They’ve already seen similar distortions in measurements of loneliness. They’re currently testing personality metrics too.
The implication is unsettling. Much of modern psychological science might be built on shaky foundations. The tools we use to define mental health don’t speak the same language to everyone.
